MiR-146a-5p Targeting SMAD4 and TRAF6 Inhibits Adipogenensis Through TGF-β and NF-κB Signal Pathways in Porcine Intramuscular Preadipocytes

Background: Intramuscular fat (IMF) content is a vital parameter to assess pork quality. Increasing evidences have shown that microRNAs (miRNAs) play an important role in regulating porcine IMF deposition. Here, a novel miRNA implicated in porcine IMF adipogenesis was found, and its effect and regulatory mechanism was further explored on the proliferation and differentiation of intramuscular preadipocytes. Results: By miRNAs sequencing analysis in porcine adipose tissues, we found that miR-146a-5p was a potential regulator of porcine IMF adipogenesis. Further study showed that miR-146a-5p mimics inhibited the proliferation of porcine intramuscular preadipocytes, but its inhibitor promoted cell proliferation. Interestingly, miR-146a-5p mimics also repressed preadipocyte differentiation, whereas its inhibitor promoted adipogenic differentiation. Mechanistically, miR-146a-5p suppressed cell proliferation by directly targeting SMAD family member 4 (SMAD4) to attenuate the TGF-β signal. Moreover, miR-146a-5p inhibited the differentiation of intramuscular preadipocytes by targeting TNF receptor associated factor 6 (TRAF6) to weaken the NF-κB signaling of the TRAF6 downstream pathway. Conclusions: MiR-146a-5p targeting SMAD4 and TRAF6 inhibited porcine intramuscular adipogenesis through attenuating TGF-β and NF-κB signals, respectively. These ndings provided a novel miRNA biomarker for regulating intramuscular adipogenesis to promote pork quality.

transcriptome sequencing technology [12]. We therefore speculated that these above miRNAs may be key regulators of porcine IMF deposition by their target genes.
The classical TGF-β/SMAD signal regulates cell proliferation, differentiation, migration and growth. SMAD4, as an important transmission medium, transduced extracellular signals including TGF-β and BMP to the nucleus [13][14][15]. Literatures indicated that SMAD4 regulated proliferation and migration of A549 cells [16], and Dihydromyricetin inhibited the proliferation of human choriocarcinoma JAR cells via downregulation SMAD4 expression level [17], and miR-224 mediated the proliferation of HCT116 cells by targeting SMAD4 [18]. Although SMAD4 has been discovered and studied in proliferation of different types of cells, it needs further explore whether SMAD4 is implicated in proliferation of porcine intramuscular preadipocytes.
TRAF6 is an important intracellular multifunctional signaling molecule, which has unique receptor binding speci city [19]. Through a series of signal transmission, TRAF6 nally activates nuclear factor NF-κB participating in cell proliferation, differentiation, embryonic development and bone metabolism [20]. At present, TRAF6 has been studied on obesity or adipogenesis. The mRNA and protein levels of TRAF6 signi cantly increased in obese mice after high-fat feeding for 6 weeks [21]. Oleanolic acid inhibited visfatin and its in ammatory response during preadipocyte differentiation by blocking IL-6/TRAF6/NF-κB signaling [22]. Furthermore, miR-345-3p regulated cell apoptosis through the TAK1/P38/NF-κB pathway by targeting TRAF6 [23]. Interestingly, miR-146 regulated various biochemical processes by targeting TRAF6, including ameliorated HFD-induced non-alcoholic steatohepatitis [24] and reducing IL-1β-induced in ammatory factors in osteoblast-derived cell lines [25]. However, it remains unknown whether TRAF6 plays a role as a miRNA target in adipogenic differentiation of intramuscular preadipocytes.
In this study, based on miRNA sequencing analysis of adipose tissue at four developmental stages of pigs at 1, 30, 90 and 240 days of age [26], we found that miR-146a-5p showed signi cant expression differences, implying it may have a key effect on IMF deposition. Our results revealed that miR-146a-5p inhibited the proliferation and differentiation of intramuscular preadipocytes by targeting SMAD4 and TRAF6 through TGF-β and NF-κB signal pathways, hinting that it is a novel key regulator in pig IMF deposition.

Animal and sample collection
Piglets (3 days) were provided by the animal experiment animal ranch of Northwest A&F University. According to the regulations of the Animal Protection Committee of Northwest A&F University, all pigs were killed in the slaughterhouse. Dissect the heart, liver, spleen, lung, kidney, dorsal longest muscle (LD) and subcutaneous white adipose tissue (SWAT), and rinse with phosphate buffered saline (PBS). The samples used for real-time quantitative PCR (RT-qPCR) were frozen and stored in liquid nitrogen.
Isolation and culture of porcine intramuscular preadipocytes After the 3 days old piglets were sacri ced, the intramuscular preadipocytes in LD were extracted. Refer to this document for the speci c operation method [27].
Transfection of mimics/inhibitor NC and miR-146a-5p mimic/inhibitor Porcine intramuscular preadipocytes were seeded in 6-well, 12-well, 24-well or 96-well plates. When detecting cell proliferation, miR-146a-5p mimics or mimics negative control (MNC) (Ribobio, China) were transfected (50 nM) when the cell density reached 50-60%. During transfection, X-tremeGENE siRNA Transfection Reagent (Roche, USA) was mixed with Opti-MEM medium (Gibco, USA) for 5 minutes, then the two mixtures were mixed for 20 minutes and added to the cell culture medium, and the medium was replaced with fresh culture after 12 hours. Cells were harvested 24 hours after transfection for cell proliferation studies. When transfected with miR-146a-5p inhibitor, the method is the same as above, but the nal concentration of miR-146a-5p inhibitor was 100 nM. For the differentiation of preadipocytes, the cells are transfected when the cell density reaches 70%. When cells grow to con uence after transfection, adipogenic differentiation begins by switching to differentiation medium.
Total RNA extraction, RNA reverse transcription and RT-qPCR After obtaining the cells, the cells were lysed with Trizol reagent (TakaRa, Otsu, Japan) and the total RNA in the cells was extracted. The concentration of total RNA was measured by the NanoDrop 2000 (Thermo, Waltham, MA, USA). Then the reverse transcription kit (TakaRa, Otsu, Japan) was used to synthesize cDNA. The speci c reverse transcription primers and procedures were used for miRNA inversion. About real-time quantitative PCR, the SYBR green kit was used and three replicates were set up, and then the PCR reaction was performed on the Bio-Rad iQTM5 system. GAPDH was used as the internal reference for all genes for standardized analysis. But when analyzing miR-146a-5p levels, U6 was used as an internal reference. Table 1 shows the primer sequences used for qPCR. The primer sequences used for qPCR were shown in Table 1.

Western blots
Cell samples were lysed using radio immunoprecipitation assay (RIPA) buffer (Beyotime, China) supplemented with protease inhibitor (Pierce, WA, USA) and total protein was extracted. The total protein samples were separated by electrophoresis in SDS-polyacrylamide gel. Then transferred it to PVDF membranes (Millipore, Bedford, MA, USA). After blocking the membrane in 5% skim milk for 2 hours, the primary antibody was incubated overnight (4 °C) and the secondary antibody was incubated for 1.5 hours (4 °C). Protein bands were exposed with chemiluminescent reagents (Millipore, Bedford, MA, USA) and quanti ed using Image J. Target prediction and luciferase activity assay The target genes of miR-146a-5p were predicted with Target-Scan 7.0. For the dual-reporter assay, we constructed a wild-type and mutant psiCHECK-2-reporter vector containing the target genes SMAD4 and TRAF6 3′ UTR region (TongYng). HEK293T was seeded in a 48-well plate and cotransfected with miRNA mimics or the negative control with psiCHECK-2-SMAD4 (or TRAF6)-reporter vector or mutant vector. After 48 h of transfection, the relative luciferase activity of Renilla compared with re y was measured.

EDU imaging assay
We used the Cell-Light™ EdU Apollo® 567 In Vitro Imaging Kit and con gured the mixed solution according to the instructions. The preadipocytes in the normal growth stage were treated with 50 µM EDU medium for 2 h. After the cells were xed with 4% paraformaldehyde, they were stained with Apollo reaction solution. Then cell nucleus was stained with Hoechst. Nikon TE2000 microscope (Nikon, Tokyo, Japan) was used to take pictures, and the data was analyzed using Image J.

Cell Counting kit 8 (CCK8) analysis
Preadipocytes were seeded to 96-well plate in a number of 4 × 10 3 cells. Preadipocytes were transfected with miR-146a-5p mimics / inhibitor or mimics / inhibitor negative control with 3 repetitions. After treatment for 24 h we switched the cells to culture medium containing 10% CCK solution for 2 h at 37 °C followed by measuring absorbance at 490 nm.

Flow cytometry
After 24 hours of cell treatment, rst wash 3 times with PBS, then x the cells with 70% alcohol at -20 °C overnight, then treat with RNaseA at 37 °C (1 mg/mL, 40 min), and stained with propidium iodide (PI) at 4 °C (50 mg/mL, 1 h). The samples were detected using a FACS Calibur ow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). The proliferation index (PI) shows the proportion of mitotic cells among the 10,000 cells examined. (4′,6-diamidino-2-phenylindole; Roche) was incubated for 10 min, then the section was rinsed with PBS.
After treatment, the sections were observed under uorescence microscope (Nikon, Tokyo, Japan).

Statistical analysis
All charts were created using GraphPad Prism 6.0 and the data represent the mean ± SEM. The signi cance of differences between the groups was assessed using the Student's t test or one-way analysis (*, P 0.05; **, p 0.01).

Results
MiR-146a-5p is a potential regulator in porcine IMF adipogenesis To screen out miRNAs related to porcine IMF deposition, miRNA-sequencing data were analyzed. As shown in the heat map, the levels of miR-146a-5p of 30 d, 90 d, and 240 d porcine adipose tissue were signi cantly higher than that of 0 d piglets (Fig. 1A). Moreover, the seed sequence of miR-146a-5p in humans, pigs and mice is highly conserved (Fig. 1B). KEGG pathway analysis predicted that miR-146a-5p was involved in the TGF-β and NF-κB pathways (Fig. 1C), GO term analysis suggested that miR-146a-5p could regulate cell proliferation and fat cell differentiation (Fig. 1D). Furthermore, miR-146a-5p highly expressed in porcine WAT (Fig. 1E). Most importantly, the levels of miR-146a-5p increased rst and then decreased in proliferated and differentiated porcine intramuscular preadipocytes, but they showed an upward trend in the late stage of differentiation ( Fig. 1F and G).

MiR-146a-5p mimics inhibits proliferation of porcine intramuscular preadipocytes
To investigate the effect of miR-146a-5p on the proliferation of porcine intramuscular preadipocytes, the miR-146a-5p mimics and mimics negative control (MNC) were transfected into cells. Compared with the MNC group, the positive cells labeled with EDU and total number of cells in the mimics group signi cantly reduced (P < 0.05) ( Fig. 2A-C). In addition, the number of S-phase cells in mimics group was signi cantly less than MNC group (P < 0.05), but the number of cells in the G1-phase was signi cantly more than MNC group (P < 0.05) ( Fig. 2D and E). Furthermore, the miR-146a-5p mimics sharply increased the level of miR-146a-5p (P < 0.05), signi cantly decreased the mRNA levels of cyclin B, cyclin D and cyclin E, whereas apparently increased the mRNA levels of P21 (P < 0.05) ( Fig. 2F and G). Similarly, the miR-146a-5p mimics downregulated the protein levels of cyclin D, cyclin E and PCNA (P < 0.05), and P21 protein tended to upregulated, but not reach statistical signi cance ( Fig. 2H and I).
MiR-146a-5p inhibitor promotes porcine intramuscular preadipocyte proliferation To validate the role of miR-146a-5p inhibitor on the proliferation of porcine intramuscular preadipocytes, inhibitor negative control (INC) and miR-146a-5p inhibitors were transfected into porcine intramuscular preadipocytes. The results indicated that knockdown of miR-146a-5p signi cantly increased the number of EDU positive cells, S -phase cells and the total cells (P < 0.05) (Fig. 3A-E). The inhibitor effectively decreased the levels of miR-146a-5p (P < 0.05), and increased the mRNA levels of cyclin B, cyclin D and cyclin E, but reducing the level of P21 (P < 0.05) ( Fig. 3F and G). Meanwhile inhibitor upregulated the protein levels of cyclin D and cyclin E, whereas downregulated the protein level of P21 (P < 0.05) ( Fig. 3H and I).
MiR-146a-5p targeting SMAD4 inhibits the proliferation of porcine intramuscular preadipocytes by TGF-β signaling pathway To further explore the regulatory mechanism of miR-146a-5p on porcine intramuscular preadipocyte proliferation, we predicted and veri ed its target genes and signaling pathway. SMAD4 may be the target gene of miR-146a-5p using TargetScan 7.0 analysis ( Fig. 4A and B). The dual-luciferase reporter (DLR) assay results showed that the relative luciferase activity of miR-146a-5p mimics plus SMAD4 WT vector co-treated group signi cantly reduced (P < 0.05) ( Fig. 4C and D). Next, the rescued experiments were further carried out. Compared with the mimics group, the number of EDU-positive cells and the total cells numbers in the mimics and SMAD4 overexpression vector co-treatment group markedly increased (P < 0.01), and rescued or even exceeded the NC group level (Fig. 4E-G). Moreover, SMAD4 overexpression restored the mRNA and protein levels of SMAD4 and cell cycle-related genes (P < 0.05) (Fig. 4H-K). In addition, the SMAD4 downstream TGF-β signaling was attenuated by mimics, but was rescued when SMAD4 overexpression (P < 0.05) (Fig. 4J and K).
MiR-146a-5p inhibitor accelerates porcine intramuscular preadipocyte differentiation To further validate the role of miR-146a-5p in the differentiation of porcine intramuscular preadipocytes, we carried out the experiments of INC and inhibitor treatment on the cells. Compared with the INC group, the lipid droplets apparently accumulated in intramuscular adipocytes, and the TG content also signi cantly increased (P < 0.05) (Fig. 6A-D). Inhibitors signi cantly decreased the level of miR-146a-5p (Fig. 6E), but markedly increased the mRNA levels of C/EBPβ, PPARγ and FABP4 (P < 0.05) (Fig. 6F).
Meanwhile the protein levels of C/EBPβ and PPARγ signi cantly increased in the treatment group (P < 0.05) ( Fig. 6G and H).
MiR-146a-5p targeting TRAF6 inhibits the differentiation of porcine intramuscular preadipocytes via NF-κB signaling pathway To further investigate the mechanism by which miR-146a-5p regulates the differentiation of intramuscular preadipocytes, we explored its target genes and signaling pathways. Online software predicted that miR-146a-5p could be combined with TRAF6 3'UTR ( Fig. 7A and B). The DLR assay result showed that the relative luciferase activity of miR-146a-5p mimics and TRAF6 WT vector co-treated group was signi cantly reduced (P < 0.05) ( Fig. 7C and D). The rescued experiments were further performed. Compared with the mimics group, the lipid droplets and TG content in the TRAF6 overexpression vector plus mimics co-treated group signi cantly increased and rescued to the NC group level (P < 0.05) (Fig. 7E-H). Moreover, TRAF6 overexpression rescued the mRNA and protein levels of TRAF6 and adipogenic related genes (P < 0.05) (Fig. 7I-L). NF-κB as an important one of TRAF6 downstream signaling molecules, the levels of NF-κB and its phosphorylation levels were signi cantly decreased in mimics group but increased in co-treated group (P < 0.01) ( Fig. 7M and N).

Discussion
As a member of the non-coding RNA family, miRNA has a crucial regulatory role in preadipocyte adipogenesis. Based on bioinformatics analysis of miRNAs sequencing data, we found that miR-146a-5p was differentially expressed during SWAT deposition in pigs. Further study showed that the miR-146a-5p sequence was highly conserved, and its function was involved in fat cell proliferation and differentiation by TGF-β and NF-κB signal pathways using KEGG and GO analysis. Moreover, miR-146a-5p not only highly expressed in porcine WAT, but also its expression levels increased rst and then decreased in proliferated and differentiated porcine intramuscular preadipocytes. Based on above analysis, we therefore speculated that miR-146a-5p also was implicated in IMF deposition.
It is a vital task to improve pork quality via control IMF content in pig production. The present study demonstrated that miR-146a-5p played a crucial role in regulating porcine IMF adipogenesis. MiR-146a-5p targeting SMAD4 inhibited the proliferation of porcine intramuscular preadipocytes through attenuating TGF-β signaling, whereas it targeting TRAF6 repressed differentiation by weakening NF-κB signaling. These ndings indicated that miR-146a-5p could be a novel negative regulator of porcine IMF deposition.
The IMF deposition depended on the proliferation and differentiation of intramuscular preadipocytes. Our results con rmed that miR-146a-5p inhibited the proliferation of intramuscular preadipocytes by reducing the number of S-phase cells and downregulating the mRNA and protein levels of cyclin B, cyclin D, cyclin E, PCNA and upregulating the mRNA and protein levels of P21. Previous studies revealed that miR-146a-5p promotes lung cancer cells proliferation by targeting claudin-12 [28], and overexpression of miR-146 or knockout of its target gene notch 1 inhibited the proliferation of mouse neural stem cells under serumfree medium [29]. Therefore, miR-146a-5p discrepantly modulated proliferation in different types of cells. Generally, a gene or miRNA has the opposite effect on cell proliferation and differentiation. Recent studies showed that miR-664-5p promoted myoblast proliferation and inhibited myoblast differentiation [30], and miR-429 accelerated proliferation of porcine preadipocytes and repressed adipogenic differentiation [31]. Interestingly, in this study, miR-146a-5p repressed both proliferation and differentiation of intramuscular preadipocytes. Literatures indicated that miR-483 inhibited both the proliferation and differentiation of bovine myoblasts [32], as well as miR-342-5p restricted the proliferation and differentiation of osteoblasts by inhibiting the expression of Bmp7 [33]. Therefore, our results were reasonable because of complexity of the biological processes of intramuscular preadipocytes via miRNA regulation.
Generally, miRNAs regulate different biological processes in the same cell by different target genes. Therefore, we predicted the target genes of miR-146a-5p in cell proliferation and adipogenic differentiation, respectively. During the proliferation phase, we predicted that SMAD4 was the target gene of miR-146a-5p. Recent studies showed that miR-145-5p inhibited the proliferation of ovarian epithelial cancer cells by targeting SMAD4 [34], and the overexpression of miR-663a suppressed the hepatic stellate cell proliferation by downregulating SMAD4 level [35]. Based on above results, SMAD4 is mostly a positive regulator of cell proliferation. Moreover, The SMAD4 and TGF-β signaling pathways played important roles in the process of miRNA regulating cell proliferation, respectively [36,37]. Here, we revealed that miR-146a-5p targeting the SMAD4 was identi ed as a novel miRNA that repressed proliferation of porcine intramuscular preadipocytes via the TGF-β signaling pathway.
Furthermore, we con rmed that the target gene of miR-146a-5p was TRAF6 during adipogenic differentiation of porcine intramuscular preadipocytes through TargetScan 7.0 analysis, luciferase activity assay and rescued experiments. TRAF6 is a signal transduction factor that connects cell surface receptors with intracellular signal proteins. In addition to the in ammatory immune response, TRAF6 also regulated cell differentiation and survival [38]. Previous studies showed that inhibiting the CD40-TRAF6 interaction treated obesity by improving glucose tolerance and reducing the accumulation of immune cells into adipose tissue [39]. Moreover, green tea extracts reduced the adipose tissue weight of obese mice by reducing the expression of TRAF6 [40]. Therefore, TRAF6 has a positive regulatory effect on fat accumulation. Importantly, the NF-κB signaling which was the TRAF6 downstream inhibited in ammatory injury, promoted chondrocyte apoptosis and repressing pancreatic cancer cell proliferation [24,41,42]. In this study, we uncovered that miR-146a-5p targeting TRAF6 inhibited differentiation of porcine intramuscular preadipocytes by NF-κB signaling pathway.

Conclusions
In conclusion, miR-146a-5p targeting SMAD4 inhibited the proliferation of porcine intramuscular preadipocytes through TGF-β signal pathway, whereas targeting TRAF6 repressed adipogenic differentiation via NF-κB signal pathway (Fig. 8). These ndings not only provide a novel miRNA biomarker to modulate IMF content for promoting pork quality, but also help us to better understand the role and regulatory mechanism of miRNAs on IMF adipogenesis.

Availability of data and materials
All data generated or analyzed during this study are available from the corresponding author by request.
Ethics approval and consent to participate        miR-146a-5p directly targets TRAF6 mRNA to inhibit its translation and inhibits adipogenesis gene expression through the NF-kB signaling pathway, thereby inhibiting adipogenesis.